Seven Mycobacterium ulcerans strains were investigated. Five were obtained from biopsies of BU patients from two BU endemic areas (Amasaman, Nsawam) in Ghana, while 2 were American Type Culture Control (ATCC) reference isolates (Ghana #970321 (D19F9), Benin #990826 (D27D14)) provided by Professor Françoise Portaels of the Institute of Tropical Medicine, Antwerp, Belgium. Prior to their use, the 2 reference and 5 clinical isolates were confirmed for their authenticity and purity as M. ulcerans isolates by phenotypic tests (WHO, 2001) and polymerise chain reaction (PCR) (WHO, 2001).

Nineteen microorganisms (6 Gram-positive bacteria, 10 Gram-negative bacteria and 3 fungi) were used to investigate any extended antimicrobial property the herbal preparations may possess. They microbes consisted of 11 biochemically-authenticated clinical isolates (Stretococcus sp., Staphylococcus sp., Bacillus sp., Haemolytic E. coli, Klebsiella sp., Salmonella sp., Salmonella typhimurium, Pseudomonas aeruginosa, Shigella sonnei, Aspergillus sp., Candida albicans) and 8 reference isolates (Staphylococcus aureus – ATCC 6538; Bacillus cereus – VDL 173; E. coli – ATCC 8739; Pseudomonas sp.- ATCC 27853; Enterococcus faecalis – VDL 122; Salmonella adabraka – VDL 275; Bacillus subtilis – ATCC 6633 and Aspergillus niger – ATCC 16404)

Forty-four medicinal plants and herbal preparations (Table 1) were obtained for the study, based on their purported antimicrobial and anti-inflammatory properties, which are exploited for the treatment of Buruli ulcer, various skin conditions, sores, new and old wounds, gastric and diabetic ulcers and cancers. The herbal preparations were obtained from: (i) Traditional herbal practitioners, (ii) a vendor of traditional herbal medicines, (iii) 2 herbal shops and (iv) users of traditional herbal preparations. All the plant materials were authenticated at the Botany Department of the University of Ghana, with the exception of 4 imported herbal preparations. The 4 imported herbal preparations (Swedish bitters, Hydrastis canadensis, Chamomilla recutita and Symphytum officinale) have extensive patronage in Ghana. Swedish bitters is an already-bottled alcoholic extract of 11 herbs (Aloe, Myrrh, Saffron, Senna, Camphor, Rhubarb root, Zedvoary, Manna, Theriac venezian, Carline thistle root, Angelica root) produced in Germany for NATUREWORKS, a division of ABKIT, INC, New York, NY 10128; while Hydrastis canadensis (dried roots), Chamomilla recutita (dried flowers) and Symphytum officinal (dried roots) are imported as such from Germany and retailed by a certified Herbal shop (Herbal Options Ltd, Ghana).

Pefloxacin mesylate dehydrate and miconazole were used as reference drugs for the Gram-positive bacteria /Gram-negative bacteria and fungi respectively. Pefloxacin mesylate dehydrate is an antibiotic produced by WORKHARDT LTD of India, while miconazole is an antifungal produced by JANSSEN-CILAG LTD of Portugal.

The study was undertaken in a level 2 biosafety laboratory. Microbiological procedures such as subculturing, inoculum preparation, standardization of microbial isolates and the inoculation of media with microbial isolates were all conducted in a class IIA/B₃ biosafety cabinet, while the technical team observed all institutional (NMIMR) biosafety guidelines for personnel protection and that of the laboratory.

The M. ulcerans isolates were subcultured on Lowenstein-Jensen (L-J) slants at 32 C for eight weeks, during which period they were examined for growth and contamination. Each of the M. ulcerans subcultures was suspended in sterile distilled water and standardized by the McFarland nephelometric standard. Briefly, a drop of sterile distilled water was added to a test tube containing 15-20 sterilized glass beads to wet them. A loopful of M. ulcerans subculture was added to the beads and drops of sterile distilled water were added intermittently and vortexed to break up the M. ulcerans colonies and to adjust the turbidity of the suspension to that of a No. 1 McFarland. Since M. ulcerans in culture is very waxy and not easily dispersible in water, all manipulations were done on ice. The No. 1 McFarland standardized M. ulcerans suspensions were serially diluted 10-fold to yield 10-2 and 10-4 suspensions. Smears of the resultant M. ulcerans suspensions were stained with Ziehl Neelsen (ZN) stain for detection of acid-fast bacilli and to check for microbiological purity. The serially diluted suspensions were later used for the susceptibility test.

The medicinal plants were prepared into juices, infusions or decoctions and incorporated into media as such, to simulate the state in which consumers would use them. Swedish bitters, being a liquid extract was incorporated into media as such; also simulating the state in which it would be used.

Fresh rhizomes, bulbs, thick leaves and plants containing gel were prepared into juices. Weighed amounts of each herb was washed in running water, rinsed with sterile distilled water and blended with sterile distilled water in a sterile Waring laboratory blender. The resultant 20% w/v juice was strained with a sterile tea strainer and filtered through a sterile Whatman No. 1 filter paper. The filtrate was kept at 4 C until use the following day.

Weighed amounts of whole plants or leaves were washed under running water, rinsed with sterile distilled water, shredded into bits, macerated overnight in sterile distilled water and subsequently boiled in a covered glass jar for 5 minutes. The infusion was strained in a sterile tea strainer, while pressing down the leaves (to get as much of the aqueous component from the plant as possible). The resultant 20% w/v infusion was left to cool and filtered through a sterile Whatman No. 1 filter paper and kept at 4 C until use the following day.

Weighed amounts of stems, barks and roots were prepared as decoctions. The stems, barks and roots were scrubbed in running water, rinsed with sterile distilled water, air dried in a clean air rack and dry blended into powder. The powders were then macerated overnight in sterile distilled water and subsequently boiled in a covered glass jar for 20 minutes. Each decoction was strained in a sterile tea strainer, while pressing down the powder (to get as much of the aqueous component from the plant as possible). The resultant 20% w/v decoction was left to cool and filtered through a sterile Whatman No. 1 filter paper and kept at 4 C until use the following day.

Twenty-five (56.82%) of the 44 herbal preparations inhibited the growth of all the 7 Mycobacterium ulcerans isolates investigated (Table 1). Five (20.00%) of the herbal preparations namely; Tonic 1, Aglaonema commutatum, Zanthoxylum xanthoxyloides , Spathodea campanulata and Gratiola officinalis are purportedly being used for the treatment of Buruli ulcer by the Traditional herbal practitioners (THPs) who provided them.

Figure 1

Table 1a: Test Results Of Herbal Preparations Investigated For Activity Against

Figure 2

Table 1b: Test Results Of Herbal Preparations Investigated For Activity Against

MICs of the 10 herbal preparations are presented in Table 2. Briefly, the MICs were between 0.20 and 12.50% vol/vol (i.e. 1:640 – 1:2 dilution), with Hydrastis canadensis exhibiting on average, the highest inhibitory activity (1.73 ± 1.38% vol/vol), followed by the white variety Allium sativum (2.40 ± 2.0% vol/vol), while Syzygium aromaticum and Spathodea campanulata exhibited the lowest inhibitory activity (12.5 ± .00% vol/vol % vol/vol). The differences in the MICs were statistically significant (P 0.01) and the eta2 coefficient (0.655 = 65.5%) also showed that the herbal preparations greatly accounted for the variations in the MIC values obtained.

Figure 3

Table 2: Minimum Inhibitory Concentrations (% Vol/Vol) Of Herbal Preparations With Antimicrobial Activity Against Isolates

Table 2 shows that all the 7 M. ulcerans isolates were susceptible to each of the herbal preparations, but at different susceptibilities. The average susceptibilities of the 7 M. ulcerans isolates were between 5.08 ± 4.31% vol/vol and 8.67 ± 5.14% vol/vol. The Amasaman #2 isolate was on average the most susceptible (5.085.08 ± 4.31% vol/vol) to the herbal preparations, while the Ghana reference isolate was the least susceptible (8.67 67 ± 5.14% vol/vol). However, the differences in susceptibilities of the 7 isolates were not statistically significant (P 0.05) and were subsequently confirmed by a low eta2 coefficient (0.050 = 5.0%).

The comparative susceptibilities of the isolates by community presented in Table 2 shows that the Amasaman group of isolates was on average (6.10 ± 4.53% vol/vol ) more susceptible to the herbal preparations than the Nsawam group of isolates (6.64 ± 5.17 %vol/vol). However, the difference in their susceptibilities was not statistically significant (P 0.05). The low eta2 coefficient (0.003 = 0.3%) further confirmed the weak effect that the source (community) of the isolates had on their susceptibilities to the herbal preparations.

All the 10 (100%) herbal preparations also demonstrated extended antimicrobial activity against some or all the fast growing microbes investigated, as shown in Table 3.

Figure 4

Table 3: Extended Antimicrobial Profile Of Buruli Ulcer Herbal Preparations

Given that 56.81% (n=44) of the herbal preparations that were selected on the basis of their purported dermatological and antimicrobial properties also demonstrated anti-M. ulcerans activity, goes to confirm the assertion by several ‘phytoscientists' that the most appropriate and surest way to identify higher plants for drug discovery is by follow-up of their ethnomedical uses (Spieler, 1981; Ogura et al., 1982; Oubre et al., 1997; Fabricant and Farnsworth, 2001). Furthermore, given that 21.74% (n=24) of the effective herbal preparations (Tonic 1, Zanthoxylum xanthoxyloides, Spathodea campunalata, Gratiola officinalis and Aglaonema commutatum) were provided as BU herbal medicines by 3 THPs, lends credence to their claims of successful treatment of BU patients with herbal medicines, while also indirectly confirming reports of successful herbal treatment of BU patients by other THPs. This is a hopeful finding because most of the effective herbal preparations are those commonly used for the treatment of sores, wounds, swellings and ulcers (of other etiological origin) in our traditional settings; suggesting that BU treatment may not be as elusive as we have come to think. The study has primarily shown that there are a number of herbal preparations in Ghana with anti-M. ulcerans activity; hopefully, the same plants may be found in other African and tropical countries with similar climatic conditions. In view of this plausible assumption, since BU is endemic not only in Ghana but other African and tropical countries (Portaels et al., 1998; WHO, 2001, Van der Werf et al, 2005), the identified herbal preparations should be investigated further since some may be effective in-vivo and consequently help address the large-scale difficulties associated with the treatment of BU. The herbal preparations that we have identified could also offer scientists who are interested in isolating bioactive compounds a wide range of choices for BU drug research and development.

The differences in the MICs of the herbs were statistically significant, suggesting that some of the herbal preparations may be more effective than others; for which reason, there is the need to decide on the herbal preparations that merit further investigation. According to van den Berghe & Vlietinck's (1991) ‘rule of thumb', a prominent plant with antibacterial effect that is worthy of further investigation, is one that apart from the 1:2 dilution, the 1:8 and 1:32 dilutions also show inhibitory activities. Against this background, 3 of the preparations namely; Hydrastis canadensis (1:8 - 1:128), Allium sativum, white variety (1:4 - 1:64) and Gratiola officinalis (1:2 - 1:32) merit further investigation without argument. However, Allium sativum, purple variety (1:8 - 1: 16), Tonic 1 (1:2 – 1:8) and Tonic 3 (1:2 – 1:8) could also be included since they are somewhat within the range. These selections notwithstanding, in-vivo results sometimes fail to support in-vitro findings, therefore there is a risk in adhering strictly to the rule because a selected herbal preparation may fail in-vivo, while a rejected herbal preparation may rather prove effective in-vivo. This is confirmed by the modest performance of Zanthoxylum xanthoxyloides, Spathodea campanulata and Tonic 1 in-vitro, while they are being used effectively in BU patients (personal communication). Anyway, the ultimate guide to treatment of any infection must be clinical outcome; against this backdrop, all the herbal preparations that have demonstrated activity in-vitro should be investigated further in-vivo for identification of ‘the' useful BU herbal preparation. In the interim, the herbal preparations that have been identified in our study could be used as poultices on the lesions, while further studies are concluded regarding the possibility of their systemic application.

All the 6 Ghanaian M. ulcerans isolates (though few) were successfully inhibited by all the 10 herbal preparations, suggesting that other Ghanaian M. ulcerans isolates may also be susceptible to the same batch of herbal preparations. Secondly, the differences in susceptibilities amongst the individual isolates on one hand and between the communities on another hand were negligible; a finding that was also confirmed by the eta squared coefficients which showed that the successful inhibition of the M. ulcerans isolates was largely due to the herbal preparations (0.655 = 65.5%) and negligibly due to the individual isolates (0.050 = 5.0%) and the communities (0.003 = 0.3%), 0.3%) from which the isolates were obtained. These findings suggest that herbal therapy should not be overlooked in the search for an effective BU treatment since it may hold some promise. An effective herbal medicine would provide a window of hope for both the accessible and affordable treatment of Buruli ulcer because herbal preparations are usually cheaper than orthodox drugs and can be prepared by the users themselves in most cases. This approach could enhance efforts at controlling the disease since those mostly affected by BU reside in areas where health facilities are far away and therefore not easily accessible. Besides, most BU patients occupy the lower rungs of the socio-economic ladder and as such have difficulty accessing the health services available, especially surgery ($710/case) (Amofa et al, 2002), which is currently the main treatment option (Thangaraj et al., 1999). Herbal therapy could be considered, provided the herbal preparations are found to be safe in pre-clinical studies. Preclinical safety studies may suffice for some of the herbal preparations since there are reports that herbal therapies are already being used; moreover, there is also an urgency to find treatment for BU since its prevalence is higher than that of tuberculosis and leprosy in some communities (Amofa et al., 2002). In the case of herbal preparations without supportive history, clinical trials would have to be conducted after the herbal preparations have been found to be effective and safe in pre-clinical studies.

Herbal therapy would be worth considering as a treatment option because at some stage of BU (in both BU patients and experimentally infected animals), a large number of bacilli are present (Portaels et al., 1998; Addo et al., 2005, 2007), suggesting that a favourable condition for the selection of resistant mutants could be created. It is for this reason that combination therapy (as practised in the treatment of TB) has been adopted in the treatment of BU (WHO, 2003; WHO, 2004; Etuaful et al., 2005). Some TB drugs have also been found to be effective to some extent in BU treatment and have therefore been recommended by W.H.O as the current antibiotic treatment for BU (WHO, 2003). Unfortunately the use of the TB drugs (ex. rifampicin and streptomycin) could lead to an increase in drug resistance, especially in undiagnosed TB patients, which could eventually result in an increase in TB cases, especially in TB/BU endemic communities. In view of this, herbal therapy should be actively pursued as a BU treatment option because a single herb is never a single compound but a group of compounds, which potentiate each other or create synergy (Stermitz et al., 2000). In other words, the use of a single herb could simulate combination therapy, which may prevent, or at worse, delay the development of antimicrobial resistance. Against this background, the herbal preparations that we have identified should be used as whole plants to preserve their ‘combination therapy' property, since a lot of plant extracts are known to fail because only the active ingredient is isolated, which leads to the loss of the synergy that is conferred by the rest of the plant components (Chaudhury, 1992; Duke et al., 1998; Stermitz et al., 2000). Several cases have been documented by Beckstrom-Sternberg & Duke (1994), where synergy was lost by using the single-ingredient approach to develop drugs from plants.

Observations in BU patients and M. ulcerans infected animals demonstrate that M. ulcerans disseminates into regional lymph nodes, visceral organs and bone marrows, with resultant osteomyelitis, (Portaels et al., 1998;, Addo et al., 2005; Addo et al, 2007), which is often secondarily infected by other microorganisms (WHO, 2001; Addo et al, 2007); while some BU conditions also become gangrenous (WHO, 2001; Addo et al., 2005) and cancerous with time (WHO, 2001). Therefore, the supplementary demonstration of broadspectrum antimicrobial activity by all the 10 herbal preparations investigated (especially against organisms with predilection for the skin such as Staphylococcus aureus, Staphylococcus spp., Pseudomonas aeruginosa, Bacillus cereus, Bacillus spp and Candida spp, Aspergillus spp) makes them desirable BU remedies. The use of a single herbal preparation with activity against both M. ulcerans and other microbes will simultaneously treat the M. ulcerans infection and some of the accompanying complications if caused by the microbes that they successfully inhibited. Above all, the identified herbal preparations could be combined if found to be synergistic, in order to obtain better and lasting performance.

In conclusion, this study has shown that there are a number of herbal preparations in Ghana with anti-M. ulcerans activity in-vitro, and a few others (provided by THPs who are using them for treating BU patients) with anti-M. ulcerans activity in humans. These findings both directly and indirectly respectively, confirm reports that herbal medicines are being used to treat BU in Ghana. Since BU is endemic in Ghana and other African countries (Portaels et al., 1998; WHO, 2001), and it is also currently without an effective and affordable treatment, the identified herbal preparations should be investigated with all seriousness since some may help address the difficulties associated with the treatment of BU; especially since they also posses extended antimicrobial activity which may simultaneously treat secondary infections that are often associated with BU. We appreciate that the isolates and the communities studied were very few, and as such a larger number of isolates and communities have to be investigated to generalize. That notwithstanding, the 10 herbal preparations have shown effectiveness at inhibiting the growth of M. ulcerans in-vitro and it has been demonstrated statistically that the successful inhibitions were largely determined by the herbal preparations and very minimally influenced by the isolates or sources of the isolates, therefore suggesting that the herbal preparations investigated may be effective against other Ghanaian M. ulcerans isolates. Above all, herbal therapy should be investigated further since the BU herbal preparations provided by traditional herbal practitioners successfully inhibited all the M.ulcerans isolates investigated, suggesting that herbal therapy holds some promise and could eventually be considered as one of the BU treatment options, especially in the face of the possible development of resistance to TB drugs; one of the current treatment options. Lastly, in view of the occasional absence of correlation between in-vitro and in-vivo findings, pre-clinical (safety/efficacy in-vivo) and clinical studies may have to be undertaken to select a suitable herbal preparation from among those identified in-vitro in our study.

The team members are grateful to the World Health Organization for funding the study; Professor Françoise Portaels for providing the reference isolates; Dr. Mensah Quianoo (Amasaman Hospital in the Greater Accra Region of Ghana) and Dr. Aninakwa (Nsawam Hospital in the Eastern Region of Ghana) for performing the surgeries and providing the biopsies and Mr. Amponsah of the Botany Department of the University of Ghana for authenticating the plants. The team members also acknowledge with gratitude the invaluable assistance of traditional herbal practitioners, vendors and users of traditional herbal medicines.

Department of Animal Experimentation, Noguchi Memorial Institute for Medical Research

World Health Organization, Geneva, Switzerland